Market Context — Why This Technology, Why Now

Industries worldwide are facing immense pressure to enhance operational efficiency, reduce downtime, and improve product quality in an increasingly competitive landscape. The push for Industry 4.0 and smart manufacturing mandates advanced analytics for real-time monitoring and predictive capabilities. This technology directly supports these trends by offering a robust solution for automating complex signal analysis, enabling companies to achieve significant cost savings and maintain high standards of reliability and safety without relying on scarce human expertise.

Key Competitive Advantages
01

Achieves High-Precision Feature Extraction: Converts conventional single-component signals into multi-dimensional quantities, significantly improving machine learning model input data quality. Could detect complex anomaly patterns with over 90% accuracy.

02

Establishes Market Leadership with High Uniqueness: Only 3 prior art documents highlight this technology's high uniqueness. Leveraging exclusivity until 2040, it has the potential to lead the market ahead of competitors.

03

Offers Versatile Applicability Across Broad Industries: Applicable to various time-varying signals such as manufacturing machine vibration data, IoT sensor data, and medical biosignals. Could contribute to solving challenges across diverse industries.

Market Opportunity
Manufacturing Industry
$3B–$4B globally (AI est.)
Leveraging IoT sensor data for predictive maintenance, real-time product quality monitoring, and production line anomaly detection could enhance productivity and reduce operational costs.
Industrial automation solution providers Smart factory technology developers Equipment monitoring system integrators
Energy and Infrastructure
$1.5B–$2.5B globally (AI est.)
This technology could be applied to power demand forecasting in smart grids, condition monitoring of wind turbine generators, and structural health diagnostics for bridges and roads, contributing to stable operations.
Smart grid technology companies Renewable energy asset managers Civil engineering and infrastructure monitoring firms
Medical and Healthcare
$1B–$2B globally (AI est.)
Applicable to early disease detection through biosignal analysis (e.g., ECG, EEG) and continuous health monitoring using vital data from wearable devices, contributing to preventive medicine.
Medical device manufacturers Digital health platform providers Wearable technology developers
IP Defensibility — Why Competitors Can't Replicate This
What This Patent Covers

This patent provides robust protection for a signal conversion system, a machine learning system, and a signal conversion program, covering these three aspects with 11 claims. The smooth prosecution process, including international examination and a limited number of prior art documents, indicates high technical uniqueness and strong claim stability, offering a solid foundation for licensees.

Competitive White Space

This patent primarily covers the signal conversion and machine learning system. Licensees could develop additional IP in specialized sensor hardware integration, novel data visualization interfaces, or specific control system applications leveraging the output without conflict.

Economic Impact
~$550K/year estimated anomaly detection cost reduction per facility (est.)
estimated ROI · USD · AI analysis
ROI Calculation Logic

In manufacturing equipment anomaly detection, assuming this technology improves the false detection rate from 20% to 5%. If annual losses from false detections (production halts, wasted maintenance) are ~$650K (AI est.), a 15% reduction (20%-5%) could lead to ~$100K/year (AI est.) in direct cost savings. Furthermore, a 20% reduction in sudden failures through high-precision predictive detection could save ~$450K (AI est.) from ~$2M (AI est.) in annual losses (repair costs, opportunity loss), totaling an estimated ~$550K/year (AI est.) in economic benefits.

Speed to Market
6× faster than in-house development
This technology's core algorithm, which converts time-varying signals into multi-dimensional quantities and supplies optimized time-series data to machine learning models, is already established. This significantly reduces the time required for conceptual design, algorithm development, and data preprocessing that an adopting company would need for in-house development. As a patented technology, it provides a validated framework, allowing immediate integration into existing systems or prototype development, estimated to shorten time-to-market by approximately 2.5 years.
Competitive Positioning

X: Analytical Accuracy with AI Integration
Y: Ease of Implementation and Scalability

Business Models & Applications
☁️ SaaS Data Analytics Platform
Offer a cloud-based data analytics service integrating this technology. Customers could upload sensor data to obtain high-precision feature extraction and anomaly detection results, securing stable revenue through a subscription model.
🔌 Embedded AI Module Licensing
License software modules implementing this technology to industrial equipment manufacturers and IoT device vendors. This could rapidly add high-precision AI-driven signal analysis capabilities to their products, enhancing competitive strength.
🤝 Data Science Consulting
Provide custom data analysis solutions using this technology to companies with specific industrial challenges. Offer end-to-end support from data collection to model building and operation, developing high-value consulting services.
Adjacent Application Opportunities
🚗 Autonomous Driving
High-Precision Sensor Data Analysis
Converts time-series data from autonomous vehicle sensors (LiDAR, radar, cameras) into multi-dimensional formats for machine learning models to more accurately perceive vehicle surroundings. Could enable early detection of unexpected road conditions or other vehicle movements, potentially improving safety by up to 15%.
🏙️ Smart Cities
Urban Infrastructure Anomaly Prediction
Analyzes data from various sensors (vibration, strain, flow, etc.) installed in urban infrastructure like bridges, tunnels, and water systems. Could enable early detection of deterioration or failure precursors, facilitating proactive maintenance and potentially reducing maintenance costs by 20%.
💰 Finance
Fraudulent Transaction & Anomaly Detection
Transforms time-series patterns from financial transaction data and user behavior logs into multi-dimensional inputs for machine learning models. Could rapidly detect fraudulent transactions or cyberattack indicators deviating from normal patterns, preventing losses and potentially improving detection rates by 30%.
Integration Roadmap — Estimated 11-Month Deployment
Phase 1: Technical Feasibility Assessment and Requirements Definition
Duration: 2 months
Assess integration potential with existing data sources (sensors, logs, etc.) and define the technology's scope and specific functional requirements. Conduct initial efficacy validation through a small-scale Proof of Concept (PoC).
Phase 2: Prototype Development and Data Integration
Duration: 4 months
Develop a prototype incorporating the technology's key modules based on defined requirements. Establish data integration interfaces with existing systems and perform machine learning model training and tuning using real-world data.
Phase 3: Production Deployment and Operational Optimization
Duration: 5 months
Proceed with system deployment into the production environment based on prototype validation results. Post-deployment, optimize system performance and maximize business value through continuous data collection and model retraining.
Technical Feasibility
This technology features a clear modular structure for signal acquisition, multi-dimensional conversion, time-series data output, and feature extraction via machine learning models, enabling software-based implementation. The patent claims for 'signal acquisition unit,' 'conversion unit,' 'output unit,' and 'feature output unit' suggest easy integration into existing data processing pipelines or machine learning platforms via API or SDK. It is compatible with general sensor and log data, requiring no significant hardware investment, and exhibits high affinity with existing IT infrastructure and IoT platforms.
Success Scenario
Implementing this technology could enable companies to automatically detect early signs of equipment failure or product quality anomalies from subtle signal changes that were previously overlooked. This could reduce sudden manufacturing line stoppages by 20% and improve annual production efficiency by an estimated 10%. Furthermore, automating inspection processes that relied on skilled technicians could reduce labor costs by 15% annually. The result is expected to be stable, high-quality product supply and significant optimization of operational costs.
Patent Record
APPLICATION NO.
特願2020-571028
REGISTRATION NO.
7286894
FILING DATE
2019/12/17
GRANT DATE
2023/05/29
EXPIRATION DATE
2039/12/17
PATENT HOLDER
国立大学法人山梨大学
Examination History
2021年05月18日
特許協力条約第３４条補正の写し提出書
2021年05月18日
条約３４条補正（職権）
2021年06月07日
手続補正書（自発・内容）
2021年08月16日
国際予備審査報告（英語）
2022年03月18日
出願審査請求書
2023年03月14日
特許査定